Vol. 151, No. 3JOURNAL OF BACTERIOLOGY, Sept. 1982, p. 1222-12290021-9193/82/091222-08$02.00/0Copyright © 1982, American Society for Microbiology

Insertion of Bacteriophage SP, into the citF Gene of Bacillussubtilis and Specialized Transduction of the ilvBC-leu genes

CATHERINE J. MACKEY AND STANLEY A. ZAHLER*Division ofBiological Sciences, Cornell University, Ithaca, New York 14853

Received 13 January 1982/Accepted 29 April 1982

We isolated a strain of Bacillus subtilis in which the SP,c2 prophage is insertedinto the citF (succinate dehydrogenase) gene. Defective specialized transducingparticles for the ilvBC-leu genes were isolated from phage-induced lysates of thislysogen. We isolated a group of phages that differ in the amount of geneticmaterial they carry from this region. Also, we incorporated mutant ilv and leualleles into the genomes of several transducing phages. Our phage collectionenables us to identify the cistron of new ilv and leu mutations by complementationanalysis. In this process we discovered a fourth leu cistron, leuD. Characteriza-tion of the phages confirmed the published gene order: ilvB-ilvC-leuA-leuC-leuB;leuD lies to the right of leuB.

Lambda specialized transducing phages haveproven to be powerful tools for the analysis ofthe Escherichia coli genome (see reference 8).These phages permit the construction of bacteri-al strains that are heterogenotic for virtually anygene of interest. The cistron relationship ofmutations and their dominant or recessive char-acter can then be determined by complementa-tion analysis in the partial diploids. Ultimately,the characterization of specialized transducingphages and their DNA provides detailed infor-mation concerning the structure and regulationof bacterial genes.The development of specialized transducing

phages for the gram-positive bacterium Bacillussubtilis 168 has lagged behind the E. coli-lambdamodel system. Recently, an in vitro method forthe construction of specialized transducingphages from the B. subtilis bacteriophages 4105and pll has been described (4, 5). In this labora-tory we are using in vivo techniques to constructspecialized transducing phages from a temperatebacteriophage of B. subtilis called SPP (12).SPP-mediated specialized transduction of mark-ers near the normal attachment site, attSPP, hasbeen reported (1, 9, 13). To expand the range ofgenes transducible by SP,, we employ strains ofB. subtilis that lack a functional prophage at-tachment site, attSPI. SPf3 infection of thesestrains produces lysogens, albeit at a low fre-quency, that carry the prophage at different sitesaround the bacterial chromosome. Phage lysatesof these lysogens contain transducing phages fornearby genes. We are developing methods thatemploy these phages to study the B. subtilisgenome (12a).

For our model system we chose to study acluster of genes involved in the biosynthesis ofisoleucine, valine, and leucine: ilvB and C andthe leu genes. Several years ago a genetic map ofthe ilvBC-leu region was published (11). In thatreport several leu mutants were assayed for theirleucine biosynthetic enzymes. The mutantswere deficient in either a-isopropylmalate(a-IPM) synthetase (leuA), IPM isomerase(leuB), or ,-IPM dehydrogenase (leuC). The leumutations are located between pheA and thegenes encoding the first and second enzymescommon to the isoleucine and valine pathways,ilvB (a-acetohydroxyacid synthetase) and ilvC(a-acetohydroxyacid isomeroreductase). Thegene order was reported to be ilvB-ilvC-leuA-leuC-leuB (11).At the time those studies were conducted,

there was no way to construct strains that wereheterogenotic for ilv or leu; thus complementa-tion analyses were impossible. Assignment ofnew ilv or leu mutations to the proper generequired enzyme assays, and the determinationof gene order relied on mapping mutations bytransformation. Moreover, there was no way todetermine whether IPM isomerase is encoded bytwo cistrons in B. subtilis as it is in Salmonellatyphimurium and E. coli (6, 10).

In this paper we report the isolation andcharacterization of an SPPi lysogen in which theprophage is inserted into the gene for succinatedehydrogenase (citF). We describe six classes oftransducing phages that we have isolated fromphage lysates of this lysogen. The phages differin the amount of genetic material they carryfrom the ilvBC-leu region. We also describe the

1222

on May 21, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

TRANSDUCTION OF B. SUBTILIS ilvBC-leu GENES 1223

TABLE 1. Bacterial strains used in this studyStrain Genotype' Source (reference)

CU240 ilvB2 leuB6 trpC2 (SP) (11)CU257 leuB16 trpC2 (SP) (11)CU258 ilvB2 leuA5 trpC2 (SP3) (11)CU267 ilvB2 leuB16 trpC2 (SPP) Laboratory collectionCU315 leuD17 trpC2 (SP,B) Laboratory collectionCU740 leuAS trpC2 (SP,B) (11)CU748 ilvBAI pheA2 trpC2 (SPI) Laboratory collectionCU1050 ade leu metBS sup-3 thr attSPI (12)CU1065 trpC2 attSPP (13)CU1316 ilvBAl metB5 (SP,B c2) (9)CU1846 ilvC9 leuA169 metBS (SP,B c2) Laboratory collectionCU1852 argA2 citD2 metBS pheA2 Laboratory collectionCU1881 argA2 citD2 metBS pheA2 citF::SP1 c2 SP, c2 infection of CU1852CU1886 argA2 pheA2 attSPI citF::SPP c2 CU1065 PBS; CU1881; Met+

selectionCU2322 ilvBAI leuD1l7 trpC2 (SPP c2) Laboratory collection

a attSPP means the strain has a normal, unoccupied SPP attachment site. Such strains are sensitive to SPIunless they carry an SPI3 prophage elsewhere.

construction of a set of phages, each of whichcarries a mutant allele for one of the ilvBC-leugenes. These phages have permitted us to identi-fy a fourth leu cistron, leuD. We demonstratethe usefulness of these phages in identifying newilv and leu mutations and in determining geneorder and cistron number by complementationanalysis.

MATERIALS AND METHODSBacteria and phage. All bacterial strains used in this

study were derived from B. subtilis 168 (Table 1; seeTables 2 and 3). Methods for routine culturing ofbacteria have been described (11). When strainsCU1852, CU1881, and CU1886 were cultured, 0.2%glucose was added to complex medium. Double lyso-gens, heterogenotic for ilv or leu, were cultured onminimal medium to maintain selection for the wild-type alleles. PBS1 phage lysates for strain construc-tions and mapping experiments were prepared andtransductions were carried out as described (11). Cit+transductants were selected for ability to grow withlactate as the sole carbon source (9). DNA-mediatedtransformations were carried out as described (11).

Bacteria were tested for SPI lysogeny by immunityto SPt3 cl, a clear-plaque mutant of SP, (13), and bybetacin production (2). Betacin is a bacteriocin-likesubstance produced by SP, lysogens and is active onnonlysogens. Most of the SP,B phages described herecontain the c2 mutation, which results in the produc-tion of a thermolabile SPt repressor (9). SP1 c2lysogens were heat-induced by a 10-min incubation ina 50°C temperature block during early logarithmicgrowth. SPI c+ lysogens were induced with mitomy-cin C as described (9). The phage lysates were assayedfor PFU by the soft-agar overlay technique, usingCU1050 as an indicator (9). SP,B transducing particleswere assayed by transduction of appropriate ilv or leuauxotrophs (9).

Isolation of abnormal SPO lysogens. CU1852 argA2citD2 metB5 pheA2 was constructed by PBS1 trans-duction. The citD2 mutation is a deletion that extendsfrom within the SPI prophage through the kauA andcitK genes (R. Z. Korman and P. S. Fink, unpublisheddata). The lack of citK (which encodes a-ketoglutaratedehydrogenase) makes cells that carry citD2 asporo-genous and unable to use lactate as sole carbonsource. The normal SP,B attachment site is nonfunc-tional in citD2 strains, that is, lysogenization occurs ata low frequency and at sites other than attSPI. Abroth culture of CU1852 was infected with SP, c2 at amultiplicity of infection of 1. The cells were dilutedand spread onto tryptose blood agar base (Difco)plates. The colonies that formed after 24 h of incuba-tion at 37°C were purified by streaking on tryptoseblood agar base plates containing SP,B antiserum. Theantiserum prevented false-positive betacin tests bypseudolysogens. Thirty-four SP1 lysogens were isolat-ed among 1,000 colonies tested. One lysogen, CU1881,carried SPP c2 inserted into the citF gene (see Re-sults).

Isolation of SP speIlied transducing phages andcomplementation testing. A putative low-frequencytransducing lysate for the ilvBC-leu genes was pre-pared by induction of CU1881. The lysate was used totransduce various SP,B lysogens that were auxotrophicfor ilv or leu. SPO lysogens were used as recipients fortwo reasons. First, the presence of SPI repressorprotein prevents bacterial lysis caused by superinfect-ing phage; second, the production of high-frequency-of-transduction (HFT) lysates requires a resident pro-phage to help package the defective transducingparticles.The transductants were purified and induced to

produce phage. HFT lysates for ilv or leu were identi-fied by the following simple spot test. Appropriaterecipient strains were grown in broth to mid-logarith-mic phase. The cultures were washed, and approxi-mately 5 x 10' cells were spread on selective minimalagar plates. Then 0.05-ml samples of the lysates to be

VOL. 151, 1982

on May 21, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

1224 MACKEY AND ZAHLER

tested were spotted onto the plates. The plates wereincubated at 37°C for 24 to 48 h. Lysates that producedheavy growth in the region of inoculation were labeledHFT for the appropriate markers. The ratios ofplaque-forming particles to transducing particles werethen determined quantitatively to confirm the spot testresults.

RESULTSScreening for a prophage insertion near the

ilvBC-leu gene cluster. We isolated 34 SP, lyso-gens of CU1852 argA2 citD2 metB5 pheA2. TheilvBC-leu gene cluster is located at position 2500on the B. subtilis chromosome map (3). Aninsertion of SPP into a secondary attachmentsite in this region would place the prophagebetween the pheA and argA genes, which arelocated at positions 245° and 2600, respectively.We reasoned that if 1 of the 34 lysogens carriedthe prophage between the argA2 and pheA2markers, then Arg+ Phe+ transductants of sucha strain would be cured of SP,B. A PBS1 phagelysate made in strain CU1065 trpC2 attSPO wasused to transduce the SPi lysogens of CU1852to arginine and phenylalanine prototrophy. SixArg+ Phe+ transductants from each of the 34crosses were tested for the presence of SPP. TheArg+ Phe+ transductants of one strain, CU1881,were all cured for SP,. The other 33 strainsretained the prophage in the cross. These resultsindicated that CU1881 carried an SP,B prophagebetween argA and pheA. Thus, CU1881 was apossible source of transducing phage of theilvBC-leu genes.

Characterization of CU1881: argA meBS citD2pheA2 citF::SP,3. We were not aware that theSP, insertion had occurred in the citF gene,giving strain CU1881 two defects in its citric acidcycle. We were therefore surprised to find thatwe could not transduce CU1881 to Cit+ by PBS1transduction. We tried to transduce CU1881 toMetB+, since we knew that citD2 is cotrans-duced to CitD+ 60% of the time with metB5. Allof the Met+ transductants of CU1881 were stillCit-. However, about 60% of them producedinfrequent revertants to Cit+. The revertantswere detected as papillae in cultures plated onCit+-selective media. Cells from the papillaewere purified and found to be sensitive to SP3.One of the MetB+ transductants capable ofreverting to Cit+ was labeled CU1886. It stillcarried SP, in the region between pheA andargA, but it no longer carried the citD2 deletion.We concluded that the SPP prophage had insert-ed into, and thereby inactivated, a cit genelocated between argA and pheA. The prophagecould occasionally excise accurately, reconsti-tuting the cit+ gene.Three cit genes are located in this region: citH

(malate dehydrogenase), citC (isocitrate dehy-

J. BACTERIOL.

drogenase), and citF (succinate dehydrogenase)(3). Strain CU1886 was assayed for the citF geneproduct in the laboratory of L. Rutberg at theKarolinska Institute, Stockholm, Sweden. Theresults showed that the enzyme activity of succi-nate dehydrogenase (la) from CU1886 is 0.008U * s-1 *mg of protein-', compared to an activi-ty of 0.22 in the wild-type control strain (L.Rutberg, personal communication). Therefore,we believe that the SP3 prophage lies within thecitF gene.

Isolation of SPO dleuD-1. A potential low-frequency-of-transduction lysate for the ilvBC-leu genes was obtained by the induction ofCU1881. This lysate was used as the donor in anSPf transduction of CU1846 iIvC9 leuA169metB5 (SP1 c2). Selection was for Ilv+ Leu+transductants. Fifteen transductants were isolat-ed, and phage lysates were prepared from all ofthem. At this time we knew of only three leugenes. We tested the lysates for transduction ofilvB2, iIvC4, leuA5, leuCJ24, and leuB6. Nine ofthe 15 lysates were HFT for all of these genes;the other 6 were not HFT for any of them.Subsequently we discovered a fourth leu gene,leuD, by complementation analysis. We willdescribe the identification of leuD after ourdiscussion of the transducing phages. After ourdiscovery of leuD, we tested the nine lysates fortheir ability to transduce 1euD117; they were allHFT for leuD.One of the strains that produced HFT lysates

for iIvB, ilvC, leuA, leuC, leuB, and leuD waslabeled CU1889. Phage lysates of CU1889 con-tain both PFU of SP3 and defective specializedtransducing units in a ratio of approximately 500to 1. The specialized transducing phage inCU1889 was named SPP dleuD-1.

Characterization of SPf3 dleuD-1 transduc-tants. Twenty Ilv+ Leu+ transductants pro-duced in an SP, dleuD-1 transduction of strainCU240 ilvB2 IeuB6 trpC2 (SP,) were picked atrandom, and phage lysates were prepared fromeach of them. Fourteen of the transductantsproduced HFT lysates for ilvBC-leu. The sixremaining transductants did not. The 14 HFTlysates contained both SPB and SPP dleuD-1phage particles.We wished to determine whether SPi dleuD-1

was integrated into the region of bacterial DNAhomology ("B" configuration) or the region ofphage DNA homology ("P" configuration) inthe 14 double lysogens. One way to distinguishthese two arrangements is based on gene linkageanalysis. If SP3 dleuD-1 were in the B configura-tion, then the ilvB+ gene carried by the phagewould be linked to pheA+ by PBS1. Similarly, ifSP, dleuD-1 were in the P configuration, thenilvB+ would be linked to metB+.PBS1 phage lysates produced in each of the 14

on May 21, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

TRANSDUCTION OF B. SUBTILIS ilvBC-leu GENES 1225

orgA

ilvC+ leuC+ leuD+ilvB+ |leuA5 |euB+

I I I I' I

pheA

ilvC+ leuC+ leuD+ilvB leuA leuB+I I I I I

i-

i - _- H

i

i -

I

l -

FIG. 1. Model of SPP dleuD-1 inserting into the B configuration. The seven parallel lines represent differentdefective transducing phage genomes that may be produced upon induction. Three of the seven carry the leuA5mutation.

double lysogens were used to transduce CU1316ilvBAJ metB5 (SP, c2) and CU748 ilvBAJ pheA2(SPP). Ilv+ Met' transductants of CU1316 andIlv+ Phe+ transductants of CU748 were select-ed. In 7 of the 14 double lysogens, ilvB+ waslinked to pheA and not to metB. In five of thedouble lysogens, ilvB+ was linked to metB andnot to pheA. Linkage of ilvB+ to both metB andpheA was detected in the two remaining lyso-gens.

Isolation of SP dkuD-1 phage particles thatcarry negative alleles for ilvBC-ku. To conductcomplementation tests, we incorporated ilv andleu mutations into the genome of SPi dleuD-1.The method we developed to accomplish thiswas based on the following assumptions. Whena double lysogen is induced to produce phage,excision by homologous recombination pro-duces an array of recombinant phage genomes.The recombinant classes present depend on thearrangement of genetic markers in the lysogen.To isolate a recombinant of SP, dleuD-1 thatcarries a particular ilv or leu mutation, the

transducing phage genome should be integratedinto the ilvBC-leu region (B configuration) of astrain that contains that particular mutation.Upon induction, some fraction of the defectivephage particles will incorporate the mutant alleleinto their genomes (Fig. 1).We isolated a recombinant of SPI dleuD-1

that carries the IeuA5 mutation by the followingmanipulations. CU740 leuA5 trpC2 (SPP) wastransduced to Leu+ with SP, dleuD-1. About100 Leu+ transductants were pooled and in-duced to produce phage. We expected thatabout two-thirds of the transductants were dou-ble lysogens and that, of those, approximatelyone-half carried the SPP dleuD-1 genome in theB configuration. The phage lysate producedfrom the pool of transductants was used totransduce CU258 ilvB2 leuAS trpC2 (SPP). Ilv+transductants were selected and scored for theirLeu phenotype. Nine of 50 Ilv+ transductantswere still Leu-. Four of the nine were inducedto see whether they were double lysogens. Twostrains gave HFT lysates for each of the i1vBC-

i

VOL. 151, 1982

on May 21, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

1226 MACKEY AND ZAHLER

SPOdLeuD-I(leuC 124)

.:

SP.8dLouD-1(LeuD-117)

SP.8dleuA-9

:L

SP,8dtauS-11

SPd LeuC-2

)/

,1

SPpd louD-I

bFIG. 2. Complementation test. A culture of the recipient strain CU2463 ilvB2 1euDlI7 metBS (SP,B c2) was

spread onto minimal agar plates containing methionine, isoleucine, and valine, but lacking leucine. About 0.05 mlof sterile SP,B lysates was placed onto the recipient lawn. The plates were scored for complementation after 36 hof incubation at 37°C. (a) Null phage lysates; (b) short phage lysates.

leu genes except leuA. One of these strains waslabeled CU2084. It carries two copies of leuAS.The defective phage carried by CU2084 wasnamed SPO dleuD-1 [leuA5]. By similar manipu-lations we have isolated recombinants of SP,BdleuD-1, each of which carries a mutant allelefor one of the ilvBC-leu genes (Table 2). We usedthese "null" phages in complementation tests todetermine in which cistron ilv or leu mutationslie (Fig. 2a).

Isolation of "short" transducing phages. Inaddition to SP,B dleuD-1, which carries all of theilvBC-leu genes, we isolated specialized trans-ducing phages that carry only some of thesegenes. To do this we transduced CU267 ilvB2leuB16 trpC2 (SPO) with a lysate from thecitF::SP,B strain CU1881. Ilv+ transductantswere selected and scored for their Leu pheno-types. Thirty-two of 100 Ilv+ transductants werestill Leu-. SP,B lysates were prepared from 20 ofthem. Eleven of the 20 did not give HFT lysatesfor ilvB. Presumably they were haploid trans-ductants, the result of replacement transduc-tions in which ilv+ alleles from the transducingphage replaced the recipients' ilvB2 alleles. Theother nine transductants gave rise to HFT ly-sates for ilvB+. Two of the nine complementedmutations in ilvB but not in iIvC. Therefore, thegenomes of the transducing phages in thoselysates included a complete ilvB gene, and ex-tended into but did not include all of ilvC; norwere any of the leu genes present. Similarly,four transducing phages were isolated that end-ed in leuA, one ended in leuC, and two ended inleuB (Fig. 3 and Table 3). After our discovery ofleuD (see below), we isolated a transducingphage that ended in leuD, using CU2322 ilvBAJ

1euD117 trpC2 (SPa c2) as a recipient. The useof the short phages in a complementation test isshown in Fig. 2b.

Identification of a fourth leu cistron, lekD. Atthe start of this work we knew of only three leugenes for the three leucine biosynthetic en-zymes. While trying to classify a leu mutation inour collection we discovered that it was comple-mented by SPI dleuD-1lIeuASI, SPO dleuD-1[IeuC124], and SP, dleuD-1[1euB6]. Thus thismutation, leu-117, was in a fourth leu comple-mentation group which we named leuD. Subse-quently, we identified two more mutations in theleuD complementation group.

TABLE 2. Bacterial strains lysogenic for nullphages

Strain Genotype'

CU2083 .. . ilvB2 leuA5 trpC2 (SP,)(SP, dleuD-1[ilvB2])

CU2084 .. . ilvB2 leuAS trpC2 (SP,)(SP,B dleuD-1(euA5])

CU2085 ........... ilvC4 1euC124 trpC2 (SPO)(SP,B dleuD-1[iUvC4])

CU2086 ........... iIvC4 leuCI24 trpC2 (SPI)(SPO dleuD-1[1euC124])

CU2087 ........... ivB2 leuB6 trpC2 (SP,)(SP,B dleuD-1[1euB6])

CU2089 ........... ilvBAI 1euD117 trpC2 (SPO)(SPP d1euD-1[1euD117J)

The null phages are derivatives of SPt c2 dleuD-1that were produced in bacteria lysogenic for SPP c+Since recombination occurs between whole phagegenomes and defective phage genomes during phagereplication (9), the null phages may carry either the c2or c+ allele.

SPPdLeuD-1(LeuA5)

SP4BdteuD-1(leuB6)

a

J. BACTERIOL.

on May 21, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

TRANSDUCTION OF B. SUBTILIS ilvBC-leu GENES 1227

argA citF::SPOc2E

ilvB ilvC leuA leuC leu B louD

FIG. 3. Short defective specialized transducing phages. SP, DNA is indicated by heavy lines. Horizontallines indicate the extent of bacterial DNA included in different transducing phages.

The short phage SP,B dleuB-11 complementsmutations in ilvB, ilvC, leuA, leuC, and leuB, butnot in leuD. SPP dleuD-1 complements all themutations complemented by SPP dleuB-11, plusmutations in leuD. Neither phage transducespheA. Based on the gene order citF-ilvB-leuB-pheA (11) and the complementation pattern ofthe short phages, we predicted that leuD liesbetween leuB and pheA. To corroborate thisevidence, we determined the gene order of ilvB,leuD, and pheA by transductional analysis. APBS1 phage lysate produced in a wild-typestrain was used to transduce CU232 trpC2 ilvB2pheA2. Ilv+ transductants were selected andscored for Phe+. Of 100 Ilv+ transductants, 65were also Phe+. Therefore, the distance be-tween ilvB2 and pheA2 in PBS1 transductionunits (1 - fraction cotransduced) is 0.35. WhenCU315 trpC2 leuD117 was used as the donor,Ilv+ Leu+ transductants were selected andscored for Phe+. Phe+ transductants were alsoselected. We obtained 114 Ilv+ Leu+ transduc-tants compared to 980 Phe+ transductants. Thisratio, 0.12, reflects the recombination distancebetween ilvB2 and leuD117. Only 6 of 114 Ilv+Leu+ transductants were also Phe+. Since ilvB2and pheA2 are only 0.35 transduction unitsapart, and ilvB2 and leuDi17 are 0.12 transduc-tion units apart, the small number of Phe+transductants suggests that a double recombina-tion event is required to generate Ilv+ Leu+Phe+ transductants. These data strongly suggestthe gene order ilvB2-leuD117-pheA2.We then mapped leuD1J7 with respect to

leuB6 and ilvB2 by transformation. DNA fromCU315 leu-117 trpC2 (SP3) was used to trans-form CU240 ilvB2 leuB6 trpC2 (SPI). Both Ilv+and Leu+ transformants were selected. Of 100Leu+ transformants, 56 were also Ilv+. Thisresult suggests the order ilvB2-1euB6-leu-117.

DISCUSSIONWe have isolated a strain of Bacillus subtilis,

CU1881, in which the SPP prophage is insertedinto the citF gene. The prophage insertion pre-vents the formation of the citF gene product,succinate dehydrogenase. Gene function can berestored upon prophage excision. This is thefirst example of an insertion of the SP, prophageinto a gene, and it suggests the use of SPI togenerate insertion mutations in B. subtilis.

SPI lysates of CU1881 contain specializedtransducing phages for the ilvBC-leu genes. Onephage, SP,B dleuD-1, carries all of the ilvBC-leugenes. In addition, SPf3 dleuD-1 carries the gerEgene (7), which lies between citF and ilvB (AnneMoir, personal communication).We have analyzed the transductants produced

by SP,B dleuD-1 and found that two classes areproduced. Haploid recombinants result from thereplacement of chromosomal DNA sequences ofthe recipient by homologous sequences carriedby the phage. We call such events replacement

TABLE 3. Bacterial strains lysogenic for shortphages

Strain Genotype

CU1889 ....... ilvC9 1euAl69 metB5 (SP3 c2)(SP, c2 dleuD-1)

CU2073 ....... ilvB2 IeuBl6 trpC2 (SPf3)(SP, c2 dilvB-10)

CU2075 ....... ivB2 IeuBl6 trpC2 (SP1)(SPI3 c2 dilvC-5)

CU2078 ....... ivB2 IeuBl6 trpC2 (SPO)(SPf3 c2 dleuA-9)

CU2080 ....... iIvB2 1euB16 trpC2 (SPO)(SPI c2 dleuC-2)

CU2092 ....... argA2 iIvB2 leuDII7 trpC2 (SPO)(SPI c2 dleuB-11)

pheA

SPpdilvB-1 0

SP,8dilvC- 5

SPodleuA-9

SP,8dleuC-2

SP,BdleuB-II

SPodleuD -I

% i i i i v I IV

n i

m~~~~~~~~~~~~~~~~~~~I

n i

m i

n

VOL. 1S1, 1982

on May 21, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

1228 MACKEY AND ZAHLER

transductions. Partial diploids are generated bythe addition of the transducing phage DNA tothe recipient's genome. We call such eventsaddition transductions. When SP, dleuD-l wasused as the donor, about two-thirds of the trans-ductants resulted from addition transductions.The site for integration of the transducing

phage genome in an addition transduction isdetermined by DNA homology. The defectivespecialized transducing phage genome can insertinto the region of bacterial homology in a nonly-sogenic recipient strain (9). SP3 dleuD-1 mayintegrate into either of two regions in a lysogenicrecipient: the ilvBC-leu region (B configuration),or the prophage in the normal attachment site (Pconfiguration). We determined the location ofthe transducing phage DNA in 14 double lyso-gens by showing linkage of the wild-type ilvBgene carried by the phage to pheA (near ilvBC-leu) or to metB (near the normal attachmentsite). We showed that SPP dleuD-1 is in the Bconfiguration in seven double lysogens and inthe P configuration in five. Our result with thetwo remaining double lysogens was surprising;the ilvB+ allele was linked to both pheA andmetB. Although we have not yet characterizedthese lysogens further, it is possible that a geneconversion event resulted in the conversion ofilvB2 to ilvB+ in these strains. Alternatively, twocopies of SP, dleuD-1 may be present, one inthe B configuration and one in P.We have described manipulations to produce

recombinants of SPP dleuD-1 that carry mutantbacterial alleles. By these manipulations wehave compiled a set of null phages. Each phagein the set carries a mutant allele for one of theilvBC-leu genes. We also isolated transducingphages from CU1881 that carry less bacterialDNA than SPP dleuD-1 does. Each phage car-ries a unique amount of genetic material fromthe ilvBC-leu region. We refer to this set as shortphages.We devised a simple complementation test

using both the null and short phages to identifythe gene within which any ilv or leu mutationlies. Figure 2, for example, shows the comple-mentation pattern for the IeuD117 mutation. Theleu mutation of the recipient strain is comple-mented by three of the four null phages (Fig.2a). The noncomplementing phage, SP, dleuD-1[leuD117], has a nonfunctional leuD gene; thusthe mutation must be in leuD. We have foundthat genetic complementation can easily be dis-tinguished from recombination because comple-mentation produces the wild-type phenotype ata much higher frequency.

In agreement with these results, the phageSPP dleuD-1 complemented the leu mutation,but the short phage, SP,B dleuB-11, did not (Fig.2b); therefore the mutation must lie in leuD. Our

collection of short phages is useful in someadditional ways. First, they provide informationconcerning gene order. For example, SPPdleuD-1 complements all leu mutations. SPPdleuB-1 complements mutations in leuA, leuC,and leuB, but not in leuD. This indicates thatleuD lies beyond (to the right of) leuA, leuC, andleuB. Our results for the gene order ilvB-ilvC-leuA-leuC-leuB-IeuD are consistent with thepublished order (11). Second, the short phagesprovide information about the relative order ofmutations within a cistron. For example, SPI3dleuC-2 carries bacterial DNA that ends withinleuB. When this phage is spotted onto variousleuB mutants, no complementation (heavygrowth) is seen. When it is spotted onto a leuB6mutant a few colonies result, but none areproduced on a leuB16 mutant. This result sug-gests that, although SPf dleuC-2 does not carrythe entire leuB gene, it does carry the wild-typeallele for leuB6; thus Leu+ recombinants can beproduced. Since no Leu+ recombinants are pro-duced with a leuB16 mutant, the endpoint ofSPPdleuC-2 is before, or very close to, leuB16. ThusleuB6 lies to the left (proximal to leuC) ofleuB16). This agrees with published mappingdata (11).

Similar complementation tests can be donewith ilv mutations. ilvB and ilvC are only two ofthe four known ilv genes. ilvA and ilvD lie just tothe left of the normal SPP attachment site. Ourlaboratory has specialized transducing phagesfor these genes, and they are included in ourcomplementation testing for ilv mutations (1).We have provided strong evidence for the

existence of a fourth leu cistron, leuD, whichlies to the right of leuB. It is interesting thatthere are also four leu cistrons in E. coli (10) andS. typhimurium (6), and the leu gene order in B.subtilis is identical to the order found in thosetwo species. Although we have not performedthe enzyme assays, leuD probably encodes oneof two polypeptide chains of IPM isomerase, asit does in the enteric bacteria.We plan to use our transducing phages for

ilvBC-leu to study control mutations that affectthe expression of these genes. We hope toconstruct a restriction endonuclease map of theregion and correlate it with the genetic map byusing the short phages. Finally, we are usingSP,B dleuD-1 as a source of DNA for subcloningthe ilvBC-leu genes into plasmid vectors.

ACKNOWLEDGMENTS

We thank the following people: Ruth Korman in this labora-tory for advice concerning the isolation of abnormal SPOlysogens; Lars Rutberg, Karolinska Institute, Stockholm,Sweden, for assaying succinate dehydrogenase activity instrain CU1886; and Ann Moir, Department of Genetics,University of Birmingham, Birmingham, England, for provid-

J. BACTERIOL.

on May 21, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

TRANSDUCTION OF B. SUBTILIS ilvBC-leu GENES 1229

ing us with the information that gerE is carried by SP,B dleuD-1. Finally, we thank R. Korman, P. Fink, R. Lipsky, and J.Odebralski in this laboratory for helpful discussions.

This work was supported by National Science Foundationgrant PCM78-24113 and by Public Health Service grant GM-28038 from the National Institutes of Health. C.M. wassupported by Public Health Service Genetics Training Grantno. GM-07617 from the National Institutes of Health.

LITERATURE CITED

1. F1nk, P. S., and S. A. Zahler. 1982. Specialized transduc-tion of the ilvD-thyB-ilvA region mediated by Bacillussubtilis bacteriophage SP13. J. Bacteriol. 150:1274-1279.

1a.Hederstedt, L., and L. Rutberg. 1980. Biosynthesis andmembrane binding of succinate dehydrogenase in Bacillussubtilis. J. Bacteriol. 144:941-951.

2. Hemphill, It. E., I. Gage, S. A. Zabler, and R. Z. Korman.1980. Prophage-mediated production of a bacteriocin-likesubstance by SP, lysogens of Bacillus subtilis. Can. J.Microbiol. 26:1328-1333.

3. Henner, D. J., and J. A. Hoch. 1980. The Bacillus subtilischromosome. Microbiol. Rev. 44:57482.

4. Ijima, T., F. Kawamura, H. Salto, and Y. Ikeda. 1980. Aspecialized transducing phage constructed from Bacillussubtilis phage +10S. Gene 9:115-126.

5. Kawamura, F., H. Salta, and Y. Ikeda. 1979. A method forconstruction of specialized transducing phage pll of Ba-cillus subtilis. Gene 5:87-91.

6. Margolin, P. 1%3. Genetic fine structure of the leucineoperon in Salmonella typhimurium. Genetics 48:441-457.

7. Moir, A., E. Lafferty, and D. A. Smith. 1979. Geneticanalysis of spore germination mutants of Bacillus subtilis168: the correlation of phenotype with map location. J.Gen. Microbiol. 111:165-180.

8. Nomura, M., E. A. Morgan, and S. R. Jaskunas. 1977.Genetics of bacterial ribosomes. Annu. Rev. Genet.11:297-347.

9. Rosenthal, R., P. A. Toye, R. Z. Korman, and S. A.Zahler. 1979. The prophage of SPI3dcitK-1, a defectivespecialized transducing phage of Bacillus subtilis. Genet-ics 92:721-739.

10. Somers, J. M., A. Amzalbag, and R. B. Mlddleton. 1973.Genetic fine structure of the leucine operon of Escherich-ia coli K-12. J. Bacteriol. 113:1268-1272.

11. Ward, J. B., and S. A. Zahler. 1973. Genetic studies ofleucine biosynthesis in Bacillus subtilis. J. Bacteriol.116:719-726.

12. Warner, F. D., G. A. Kitos, M. P. Romano, and H. E.Hemphill. 1977. Characterization of SP3: a temperatebacteriophage of Bacillus subtilis 168M. Can. J. Micro-biol. 23:45-51.

12a.Zahler, S. A. 1982. Specialized transduction in Bacillussubtilis, p. 269-305. In D. Dubnau (ed.), The molecularbiology of the bacilli. Academic Press, Inc., New York.

13. Zahler, S. A., R. Z. Korman, R. Rosenthal, and H. E.Hemphill. 1977. Bacillus subtilis bacteriophage SP,: lo-calization of the prophage attachment site, and special-ized transduction. J. Bacteriol. 129:556-558.

VOL. 151, 1982

on May 21, 2018 by guest

http://jb.asm.org/

Dow

nloaded from